CN100458793C - Mapping conversion method between data access level Xml format data and relational data - Google Patents

Abstract

This invention provides a mapping converting method for format data and relation data of a data access layer Xml including two periods of designing time and operation time, in which, at the designing time period, a model is set up based on the data structure of the relation data and mapping and conversion of the object and the relation data to be described as a data object and data model, and the data object means that the relation data sets up a model and abstracts the data structure of the list and view in the database to describe its name and field attribute, at the operation period, the data access an engine to structure SQL sentences dynamically for accessing database according to the element data information defined in designing and input data information to finish function of data access layer of increasing, deleting, correcting and checking automatically.

Description

Mapping conversion method between data in Xml format and relational data in data access layer

1. Technical field

The invention designs a computer application technology, and in particular relates to a mapping conversion method between Xml format data and relational data in a data access layer.

2. Background technology

The software architecture can usually be divided into three layers: interface presentation layer, business logic layer, and data access layer. Usually in the enterprise application development environment, the object-oriented development method has become the mainstream method. Objects only exist in memory. If you want to save object data permanently, you need to persist the object, that is, store the object in the database. Conversely, data needs to be loaded from the database and constructed as memory objects. At present, the relational database is still the mainstream and the most widely used database, and the form of its data storage is non-object-oriented. Therefore, to access objects with relational databases, mapping and conversion are required, which is what we usually call data access layer development. Manually writing and maintaining the mapping and conversion code of the data access layer is undoubtedly a tedious and time-consuming task due to the differences between compatible database types (Oracle, Sq1 Server, DB2, Sybase, etc.) and versions, etc.

In addition, for data in business application systems, due to reasons such as efficiency and space, one type of data often needs to define multiple entity object types, and the definition, conversion, deployment, and maintenance of these object types are also relatively cumbersome. Therefore, a common unified data format and object form are required to encapsulate the data.

3. Contents of the invention

The purpose of the present invention is to provide a mapping and conversion method between data access layer Xml format data and relational data, the method mainly includes two phases during design and runtime, wherein:

In the design-time stage, as shown in Figure 1, the data structure of relational data and the mapping and conversion of objects and relational data are modeled, described as two kinds of metadata, data objects and data models, and data objects are relational data for modeling , to abstract the data structure of tables and views in the database, and describe their names, field attributes, etc.

Among them, the table or view describes its display name, encoding, table or view name (database identifier), author, version, creation date, etc.

Among the field properties, the main information includes: field display name, field name (database identifier), field type, field length, whether it is required, whether it is a keyword, etc.

The field type in the field attribute considers that it needs to be compatible with multiple types of databases, abstracts the data type, defines a set of neutral data types, and establishes a mapping relationship with a specific database data type, thereby shielding Database differences. On this basis, some business data types commonly used in the business system are defined, such as number, name, note, amount and other data types, which have two functions: one is to facilitate input, and the other is to specify the type of business data And unification, to avoid program errors caused by manual setting errors. For business data types, extended definitions are provided, and users can expand by themselves during development and use. In addition, for the convenience of entry and maintenance, a conversion function between the document format and the database design tool PD (Power Designer) is provided. Users can design the database in PD, save it as a PD document, and directly import and generate data objects.

data model

A business entity is an organic form of data in a business system. It is generally composed of multiple data tables, and the master-slave and association relationships are formed between multiple tables. The data model is used to describe the data composition relationship of business entities. On the basis of data objects, model business entities and the relationship between business entities, describe the data structure of business entities, business data information of business entities, and the relationship between business entities.

1) The structure of the data model is: a master table, multiple slave tables, forming a master-slave relationship, and slave tables can be mounted under the slave table to form a nested master-slave relationship. A model can define relationships with other models.

2) The relationship between data models and data objects. The table in the data model is referenced from the data object, and the fields in the table in the data model correspond to the fields in the data object. This correspondence can be a complete correspondence or a partial correspondence. That is, some fields in the corresponding data object that the table of the data model can select. After corresponding, the description information of the data model is automatically formed according to the description information of the data object. The code and name of the table in the data model are directly taken from the code and name of the data object; the field description information of the table in the data model is partly taken from the data The information described by the fields of the object: data type, whether it is required, primary key, etc. In particular, the data type defined on the data model is different from the data type defined on the data object. The data type defined on the data model is the data type oriented to the development programming language, while the data type on the data object is the data abstracted from the database. type. For example, the data type "VARCHAR(n)" defined by the data object corresponds to the data type "string" in the data model.

3) Table attribute information of the data model: number, name, primary key field designation, primary key generation rules, whether to enable coded fields, coded field designations, coded field generation rules, data classification definition, sorting definition, etc.

4) Field attribute information of the data model. Field display name, number, label (referring to name), field data purpose, field type, field object type definition, whether it is used as column permission assignment, data permission definition, whether it is required, whether it is used when inserting, whether it is used when updating use etc. Some field attribute definitions are explained as follows:

Among them, the purpose of field data is: list, card, print, reference and query conditions, which are multiple choices.

Among them, the field object type definition is used to specify the association relationship of the field or the enumeration value definition used. Relationship definition is used to define the relationship between data models. The enumerated values are defined for simple entry help. Such as "yes", "no" or "male", "female" and so on. Among them, the column permission and data permission definition are used for the association definition with the permission system.

5) Definition of the master-slave relationship of the tables of the data model. Specify the association relationship between the master table and the slave table, and specify one or more fields of the slave table as foreign keys to associate with one or more fields of the master table.

6) Association relationship between data models. The association relationship between data models is composed of a field in a data model associated with a field in another data model. There is a field object type definition in the field attribute, and when it is selected as an association relationship, it can be defined in another model's association. In the association, specify the fields associated with it in the association model, the additional conditions of the association, and specify the fields of the related association model to be queried in the model, such as a customer data model, which has a region field, the associated region data model, When linking, specify the link to bring out the area code, name and other information of the area data model.

The data objects and data models defined at design time are described by a set of custom XML tags, and the storage format itself is XML document format, which is managed as metadata (data describing data) resources.

runtime phase

As shown in Figure 2-3, at the runtime stage, the data access engine, based on the metadata information defined at design time, combined with the input data information, dynamically constructs SQL statements for database access access, accesses the database, and automatically completes additions, deletions, and modifications Check the function of the data access layer. In particular, the results of the dynamic construction process are moderately cached for efficiency.

In order to be compatible with database differences, when the data access engine constructs SQL and accesses the database, according to the relevant information of the metadata, according to the type and version of the currently accessed database, and for the dialect differences between the database type and version, use the strategy mode to dynamically adjust SQL statements to suit different database types and versions.

In addition, it provides extended interface definitions for number generation rules and permission integration in business systems to use different application scenarios. In the process of development and implementation, according to the needs of different business scenarios, the extension interface can be customized and registered in the data access engine to realize the extended functions. By default, the data access engine integrates common number generation rules and the implementation of permission integration.

For the cumbersome definition and maintenance of object types, the data exchange standard of the data access engine is XML. While returning XML, it provides the function of returning the XML Schema of XML. The data access engine provides the default Xml access object, and can also use the XML access object provided by other system basic class libraries for access. And it provides good support for exchanging data and data standards between heterogeneous systems.

From the aspect of external call interface, two clients are defined: local call and We Inspur Group Shandong General Software Co., Ltd. service call, which is convenient for calling in different application development modes.

In this way, the mapping and conversion of XML-RelationDataMapping between data in the data access layer Xml format and relational data is completed, which is referred to as XR-Mapping.

The beneficial effects of the present invention are: the mapping and conversion technology between XML format data and relational data can effectively encapsulate

Access to relational databases shields the development complexity of the underlying database due to differences in database types and versions, and provides a standard XML data access and transmission format, enabling developers to achieve persistence without writing any code The access function of the persistence layer can effectively reduce the development difficulty and development risk of the persistence layer of the software system. In addition, this method simplifies the structural adjustment of business data. When the database structure and business document structure are adjusted, adjustments can be made through graphical data objects and data model designers without compilation and physical release. Complete the adjustment work. It effectively reduces the cost and risk in the process of software system implementation and maintenance, as well as the development and implementation costs of software manufacturers. Enable the software system to quickly adapt to the individual needs of users and meet the long-term goals of users' continuous development and improvement of information systems.

Description of drawings

Figure 1 is a schematic diagram of data modeling in the design-time phase;

Figure 2 is a schematic diagram of mapping and conversion between data;

Figure 3 is a high-level design diagram of the data access engine.

4. Implementation method:

The implementation method is also divided into two stages: design time (data modeling) and operation time (data access, conversion). At design time, through data modeling, the static structure of business data is described hierarchically (data object describes the database structure, and data model describes the business data structure); at runtime, according to the design-time description, dynamic analysis is completed to complete the XML data and Mapping and transformation between database relational data.

4.1 Design time

As shown in Figure 1, when designing:

(1) According to the data structure abstracted from business requirements, establish data objects (description of the table structure in the database), describe the database name of the table, the column to which the table belongs, and the relevant field information of the column (field name, type, length, precision, unique, nullable, etc.).

(2) Based on business documents abstracted from business requirements, establish a data model and describe the composition of business documents (a general business document is composed of a master table and multiple slave tables). specific:

Select a data object corresponding to the main table of the business document as the main object of the data model.

若该业务单据有从表，则在数据模型的主对象下选择各子表对应的数据对象作为数据模型的子对象。并指定子对象和主对象的关联(从表中的主表标识字段和主表标识的关联)。

If the business document has sub-tables, select the data objects corresponding to each sub-table under the main object of the data model as the sub-objects of the data model. And specify the association between the sub-object and the main object (the association between the main table identification field in the slave table and the main table identification).

分别指定主对象和子对象的主键字段、编码字段，若有数据分级信息，则需启用分级选项，指定相应的分级信息。

Specify the primary key field and code field of the main object and sub-object respectively. If there is data classification information, you need to enable the classification option and specify the corresponding classification information.

分别指定主对象和子对象字段的相关属性：

Specify the relevant properties of the main object and sub-object fields separately:

■Basic properties of fields, such as field name, field label, field description, field type, etc.

If the field is a foreign key, specify the type of the field as "association", and specify its associated "data model", define the relation

■ link related information.

■A value domain enumeration can be specified for a field.

■ Assign fields their purpose (card, list, print, etc.).

此外，可为主、子对象指定各种约束(数据删除外键检查、唯一性检查等)。

In addition, various constraints (data deletion foreign key checks, uniqueness checks, etc.) can be specified for master and child objects.

4.2 Runtime

As shown in Figure 2, the "data access engine" receives external add, delete, modify, and search requests during runtime, and completes the mapping and conversion of corresponding XML data and relational data. Requests to the "data access engine" can be divided into two categories, one is data acquisition, that is, obtaining relational data from a relational database, and converting the result into XML data; the other is data input, that is, inputting XML data and converting it into Relational data, to synchronize (insert, update) input to the database.

(1) Data acquisition, according to the input "data model" identification and query conditions, the "data access engine" dynamically forms SQL for querying the relational database according to the description of the corresponding "data model" and "data object", and obtains data results Set, and then according to the corresponding "data model" structure, constructed as XML data and returned.

(2) Data input, according to the input "data model" identification and XML, the "data access engine" dynamically constructs a relational data result set, and according to the description of the corresponding "data model" and "data object", dynamically forms a relational database Insert, update, delete SQL, execute SQL, and complete the request. When executing multiple SQLs that affect the state of the database, the "data access engine" will automatically start the database transaction.

In this way, the mapping and conversion between XML data and relational database data is completed through runtime modeling and runtime parsing.

5. Embodiment:

The present invention has been successfully applied to the Inspur GSP application development platform, and its architecture is shown in Figure 3. System designers describe the model of business data through graphical data objects and data model designers, and apply the "data model" to each module of the platform that requires relational database persistence. These modules use XML data, and complete the access operation of XML data to the relational database by calling the data access engine.

The mapping and conversion technology between XML format data and relational data effectively encapsulates the access to relational databases, shields the development complexity of underlying databases due to differences in database types and versions, and provides a standard The XML data access and transmission format enables developers to realize the access function of the persistence layer without writing any code, thus effectively reducing the development difficulty and risk of the persistence layer of the software system. In addition, this method simplifies the structural adjustment of business data. When the database structure and business document structure are adjusted, adjustments can be made through graphical data objects and data model designers without compilation and physical release. Complete the adjustment work. It effectively reduces the cost and risk in the process of software system implementation and maintenance, as well as the development and implementation costs of software manufacturers. Enable the software system to quickly adapt to the individual needs of users and meet the long-term goals of users' continuous development and improvement of information systems.

Claims (1)

1. the mapping conversion method between data access layer xml formatted data and relation data, two stages when it is characterized in that setting up separately timing and operation, wherein:

Stage when design, carry out modeling at the data structure of relation data and the mapping and the conversion of object and relation data, be described as two kinds of metadata of data object and data model; Wherein:

Data model: be to be used for relation data is carried out modeling, come out the data structure of table in the database and view is abstract, describe its title, field attribute, wherein table or view are described its display Name, coding, table or view name or Database Identification, author, version, date created; Wherein in the field attribute, main information has: the field display Name, field name, Database Identification, field type, field length, whether must fill out, whether be key word; Field type in field attribute consider to need compatible polytype database, the data type has been carried out abstract, and data type that definition one cover is neutral and concrete asking of certain data of database type are set up mapping relations, thus shadow data storehouse difference;

Data model: be the data constituent relation that is used to describe Business Entity, Business Entity is organic form of the composition of data in the operation system, generally constitute by many tables of data, constitute master slave relation and incidence relation between many tables, on the basis of data object, be used for the relation of Business Entity and Business Entity is carried out data structure, the service data information of Business Entity and the relation of asking of Business Entity of modeling and description Business Entity; Wherein:

1) the formation structure of data model is: a master meter, and a plurality of from table, constitute master slave relation, from showing down to articulate again from table, the master slave relation of nesting, the incidence relation of a model definition and other models;

2) relation of data model and data object is: the table in the data model, quote from data object, field in the field corresponding data object of showing in the data model, this corresponding relation is corresponding fully or the part correspondence, i.e. part field in the corresponding data object of the table of data model selection; After correspondence is come, automatically according to descriptive information of data object, form the descriptor of data model, the coding of the table of data model, title are directly taken from the coding and the title of data object; The field description information of showing in the data model, part is taken from the information of the field description of data object: whether data type must fill out, major key, the data type that defines on the data model is different from the data type that defines on the data object, the data type that defines on the data model is the data type towards the exploitation programming language, and the data type on the data object then is abstract in the data of database type; Such as, " this data type of VARCHAR (n), corresponding " string " this data type in data model of data object definition;

3) the Table Properties information of data model is: numbering, title, major key field appointment, major key create-rule, whether enable code field, code field appointment, code field create-rule, data staging definition, ordering definition:

4) the field attribute information of data model is: the display Name of field, numbering, label or refer to name, field data purposes, field type, the definition of field object type, whether as row right assignment, data permission definition, whether must, whether when inserting, use, whether when upgrading, use; Part field attribute definition is explained as follows:

Wherein the field data purposes is: tabulation, card, printing, reference and querying condition;

Wherein the definition of field object type is used to specify the incidence relation of field or the enumerated value definition of use: the incidence relation definition is used for the incidence relation between the definition of data model, the enumerated value definition is used for simple typing and helps, and wherein row authority, data permission define the related definition that is used for permission system;

5) master slave relation of the table of data model definition: specify master meter and, specify from one or more fields of showing related as corresponding one or more fields of external key and master meter from the incidence relation of asking of table;

6) incidence relation between the data model: the incidence relation between the data model is made of a field in related another data model of a field of data model, a field object type definition is arranged in field attribute, when being chosen as incidence relation, the model that defines this field place is related with another model, in association, specify field related with it in the correlation model, related subsidiary condition, and specify in the model field of the related relevant correlation model of inquiry, a customer data model, a regional field is wherein arranged, related area data model, specified associations is taken the area code of area data model out of in association, the information of title;

Data object that defines during design and data model are described with a self-defining cover XML tag, and it preserves the XML document form of form own, manages as the data resource of metadata or data of description;

Stage when operation, by data accessing engine, the metadata information of definition during according to design, data message in conjunction with input, the SQL statement of dynamic constructs database access visit, accessing database is finished the function that additions and deletions change the data access layer of looking into automatically, in order to raise the efficiency, the result of dynamic construction process is carried out an amount of buffer memory;

Difference because of the compatible data storehouse, when data accessing engine constructing SQL and accessing database according to the relevant information of data, type of database and version according to current accessed, difference at dialect between type of database and the version, the usage policy pattern is adjusted the XQL statement dynamically to adapt to different type of database and version;

For number create-rule in the operation system, authority is integrated, provides expansion interface definition, to use different application scenarioss; In the sharp implementation process of exploitation, according to the needs of different business scene, self-defined realization expansion interface also is registered in the data accessing engine, to realize the function of expansion; Integrated numbering create-rule and the integrated realization of authority commonly used in the data accessing engine of acquiescence;

At object type definition and safeguard loaded down with trivial details, data accessing engine data exchange standard be XML, when returning XML, the function of the XML Schema that returns XML is provided, data accessing engine provides the Xml access object of acquiescence, or the XML visit that the system-based class libraries that utilizes other provides conducts interviews to edge, and provides support for interaction data between heterogeneous system and data standard; From external call interface aspect, having defined this locality calls and two kinds of clients of we LANG CHAO GENERASOFT Co Ltd service call, be convenient to calling of different application and development patterns, so just finish mapping and conversion XML-RelationData Mapping between data access layer Xml formatted data and relation data, abbreviated XR-Mapping as.

Images